| Tai Chi Zhuang is one of the fundamental exercises of Tai Chi Quan.Practising Tai Chi Zhuang can improve participants’ fitness effectively.The postures of Tai Chi Zhuang seem pretty simple.However,participants,especially beginners,often have deviations in their understanding of the essentials.If they cannot correct their postures in time from assessment feedback,they may not benefice from the exercise efficiently.These deviations are even harmful to the physical body.While practicing Tai Chi Zhuang,plantar soles are the body parts of participants contacting the external environment directly.Therefore,the plantar pressure can reflect the participants’ body circumstances,such as functions and structure of the feet,posture control,etc.Several studies have demonstrated that classical Tai Chi movements have distinct plantar pressure characteristics.However,it is not clear what kind of dynamic features of plantar pressure involve in practicing Tai Chi Zhuang.Moreover,existing research methods lack quantitative metrics to assess the Tai Chi Zhuang training effect as time changes.To explore the dynamic features of plantar pressure and quantify metrics for evaluating the training efficiency of Tai Chi Zhuang,we design an experimental paradigm to collect data and analyze the dynamic characteristics of plantar pressure while participants practice Tai Chi Zhuang.Then we evaluate participants’ exercising efficiency dynamically as time changes.The main research contributions of this thesis are listed as follows:(1)Design a data collection and pre-processing method for the plantar pressure experiment of Tai Chi Zhuang.Accounting to research requirements,an experimental site with AILI intelligent AL-600 detector system is built to collect plantar pressure data.As a result,we have accomplished a 6-month-period data collection and a preprocessing total of eighty-four times,including six experts and thirteen participants’data.In this thesis,a hybrid filtering algorithm with a window function is adopted to eliminate the signal noise in the pressure image.Furthermore,we design an automatic partition algorithm based on the anatomy of the human foot to divide big data into six groups according to the sectors of the foot.(2)Analyze dynamic characteristics for experts’ plantar pressure.We construct the standard plantar pressure template for experts by performing time-series statistical analysis and frequency-domain Welch Power Spectrum analysis on experts’ plantar pressure distributions(pressure ratio by foot sectors,contact area ratio,and pressure intensity ratio)parameters and Center of Pressure(COP)parameters.Then,we analyze the dynamic adjustment pattern of the plantar pressure by using wavelet-transformbased multi-time scale analysis.The results show that the plantar pressure keeps adjusting dynamically when experts practice Tai Chi Zhuang.This adjustment is a slow and periodic gradual process as experts’ standing time increases.(3)Evaluate participants’ training efficiency and balance function improvement.We first compare participants’ plantar pressure distribution(both at the beginning of the experiment and the end of the experiment)parameters and COP parameters to experts’ corresponding parameters with statistical analysis.Then,we calculate the distance between the participants’ pressure ratio and the experts’ template corresponding pressure ratio with the dynamic time warping(DTW)algorithm.Next,change weights of plantar pressure are calculated with entropy weight.Finally,the sum of DTW distance by corresponding foot sector’s weight is used as the evaluation metric.The experimental results indicate that,after six-month training,the parameters gap between participants and experts is gradually reduced.In addition,training efficiency has improved as the more effect practicing time.The balance ability test results show that a youth trained with Tai Chi Zhuang regularly improves his balance ability significantly for standing on two feet and standing on one foot with eyes open compared to those who have no specific exercise habits. |